Raphaël Cancé, Barbara Castelletti-Font, Marion Cochard, Stefan Gebauer, William Honvo, Olivier Hubert, Yannick Kalantzis, Guy Levy-Rueff, Emilie Lor-Lhommet, Jean-François Ouvrard, Simon Perillaud, Laura Roccella, Béatrice Rouvreau, David Sabes, Katja Schmidt, Camille Thubin, Youssef Ulgazi, Paul Vertier, Thao Vu
In order to contribute to the national and European economic debate, the Banque de France periodically publishes macroeconomic forecasts for France, constructed as part of the Eurosystem projection exercise and covering the current and two forthcoming years. Each publication comprises a summary of the projections. Some also include an in-depth analysis of the results, along with focus articles on topics of interest.
• After a steep drop in activity in the second quarter, during the first lockdown, then a very marked rebound from June to September, the French economy is undergoing another negative shock at the end of the year linked to the resurgence of the epidemic and to the health measures put in place. This second lockdown, which was eased at the end of November with the reopening of non-essential shops, is having a significant impact, but one that is much less severe than in the spring. GDP is thus projected to decline by around 9% over the full year 2020.
• At the start of 2021, economic activity should be penalised by continuing constraints on household consumption as the health measures are expected to be lifted gradually. Our baseline scenario assumes that the epidemic will not end immediately and that vaccines will only be fully rolled out towards the end of 2021. Under these conditions, activity is only projected to return to its 2019 level in mid-2022, and the recovery will be spread over 2021 and 2022 with GDP growth of around 5% in both of those years. In 2023, growth is projected to remain slightly above 2%, which is still a high rate, albeit less unusual.
• This baseline scenario is subject to significant uncertainty and consequently we have chosen to publish only rounded figures for our projections. In addition, based on different assumptions about how the pandemic will evolve, we have prepared two alternative scenarios: a mild one where the health situation improves significantly from the start of 2021, and a severe one where the virus continues to circulate with a high degree of virulence in France and the rest of the world in 2021 and 2022. In 2021 in particular, GDP would rebound more markedly under our mild scenario (+7%), but would not recover under our severe scenario (–1%).
• In comparison, under the baseline scenario, the euro area is expected to follow a fairly similar trajectory but with a more moderate decline in GDP in 2020 (around –7%), offset by a weaker rebound in 2021 and 2022 (around +4%).
• Thanks to the cushioning provided by public finances, household purchasing power should be preserved in average terms in 2020 and 2021, despite the recession. The saving ratio should remain high in 2021 under our baseline scenario, after hitting a record of 22% in 2020. Business investment should rebound sharply in 2021 after plunging by 10% in 2020.
• Although the short-time work schemes have helped to limit the deterioration in the labour market in the short run, a delayed downturn is expected to take hold in the coming quarters, pushing the unemployment rate to a peak of close to 11% in the first half of 2021. It should then fall back towards 9% by the end of 2022.
• After weakening gradually since the start of the year, annual headline HICP inflation is projected to average 0.5% in 2020. Inflation should then pick up again very gradually, reaching just over 1% at the end of 2023.
• The government deficit should only narrow gradually after the sharp rise in 2020. In the absence of further fiscal consolidation measures, it is still expected to be close to 4% of GDP in 2023, while government debt is projected to rise to near 120% over the same horizon.
The economic rebound in the summer and at the start of the autumn 2020 was very strong, but has been temporarily halted by the resurgence of the epidemic and the new health-related restrictions put in place since October
As anticipated in our September macroeconomic projections, the third quarter of 2020 was marked by a strong rebound in activity after the shock of the lockdown in the first half (see Chart 1). GDP increased by more than 18% in the third quarter of 2020 (after –6% and –14% in the first and second quarters respectively), bringing the overall level to 4% below that observed in the fourth quarter of 2019. The size of the rebound, which turned out to be similar to that foreseen in our June mild scenario, came as good news.
According to the latest available economic data, the resurgence of the epidemic as of September and the gradual introduction of new restrictions (see Chart 2) have brought this trend to an abrupt halt. October was marked by the implementation of a curfew in the Ile de France region and eight other cities on the 17th, followed by its extension to 54 departments on the 24th, and finally the introduction of a nationwide lockdown as of 30 October.
In November, the loss of activity linked to the second lockdown is expected to have been considerably smaller than during the first lockdown. This smaller impact is the result of more marked differences in the effect of the restrictions across sectors: manufacturing, construction, certain business services and non-market services have been less affected than in the spring. Our baseline forecast, which was finalised on 25 November, takes into account the gradual lifting of the lockdown measures as of end-November. It incorporates the assumption that the loss of activity will remain significant but will shrink in December, which is in line with the results of the business surveys conducted at the end of November and start of December, and published on 14 December (for further details, see Update on business conditions in France at the start of December 2020). This renewed drop in activity should bring the overall decline in GDP for 2020 to around 9% in average terms.
This projection was finalised on the basis of the quarterly national accounts published by INSEE at the end of October. The detailed accounts published on 27 November, in which the third quarter was revised upwards slightly, have no significant impact on our projection.
In 2021 and 2022, the gradual improvement in the health situation should lead to a marked rebound in the French economy, bringing GDP back to its end-2019 level by mid-2022 in our baseline scenario
From the start of 2021 onwards, the baseline scenario incorporates a number of assumptions.
First, the assumptions regarding the international and financial environment, for which the cut-off date is 18 November (see Table A in the appendix), are the same as those used in the Eurosystem projections (published on 10 December) for the entire euro area. This environment is on the whole more favourable than in our September publication, thanks to the recent fall in oil prices, a slight depreciation in the euro effective exchange rate and a marginally more positive outlook for foreign demand for French goods and services. The latter is expected to drop sharply in 2020 (–11%), but should then see a strong rebound in 2021 (+7%).
Second, with regard to the evolution of the health situation, we assume that the lockdown measures implemented in November will bring the spread of the epidemic under control by the end of 2020. After that, however, the restrictions will only be lifted gradually in order to contain the epidemic and avoid another lockdown. In line with the public health scenario used in the Eurosystem projections for the euro area, we assume that medical solutions such as vaccines will start to be deployed in early 2021, but only gradually, and that full coverage will not be achieved until end-2021/early 2022.
Against this backdrop, the French economic recovery is projected to be dynamic in 2021 (close to 5% growth in annual average terms), but weaker than predicted in our September publication (+7.4%). In the first quarter of 2021, GDP should rebound by nearly 2% after the steep fall seen in the fourth quarter of 2020, but should still be nearly 7% below its end-2019 level (see Chart 1).
Assuming a comparable scenario for the epidemic, economic activity in neighbouring euro area countries should follow a similar trajectory to that envisaged for France (see Chart 4). Indeed, the majority of euro area countries have also experienced a second wave and have responded by bringing in new health restrictions. As a result, the rebound in euro area GDP that began in the third quarter should be interrupted by a renewed decline in the fourth quarter, and activity should only start to recover gradually over the course of 2021. Although the euro area is expected to see a smaller drop in GDP over 2020 than France (growth of around –7%), the rebound anticipated for 2021 and 2022 (around +4%) should also be smaller than in France.
Given that this projection is closely linked to the evolution of the health situation, we have also set out two alternative scenarios in the box below. In our “mild” scenario, the health situation improves more rapidly in 2021 with the widespread roll-out of medical solutions from the start of the year, and activity returns to normal faster than in our baseline scenario. In our “severe” scenario, the evolution of the epidemic forces the government to keep the restrictions in place for longer, weighing on economic activity (see Chart 3).
In our baseline scenario, the rebound in French activity in 2021 is expected to be buoyed not just by a recovery in household consumption, but also by an upswing in public demand (see Chart 5). Public sector activity is projected to return to normal after declining markedly during the first lockdown, and government investment is also expected to rebound sharply, fuelled notably by the government recovery plan. The recovery plan measures for households and businesses, in the form of subsidies and transfers, should also help to strengthen the rebound in private investment in 2021. The cuts to taxes should boost corporate margins and very gradually affect activity. On the downside, however, lasting inventory reductions by firms should weigh on the GDP recovery, especially in 2021.
The recovery in activity should continue over 2022, with growth expected to remain dynamic at 5%, driven largely by a normalisation of private consumption as the lockdown measures are lifted. The recovery plan should also continue to support private and government investment in 2022. Activity should thus return to its pre-crisis level in mid-2022, which is slightly later than predicted in our September forecast. It is expected to grow at a pace closer to potential over 2023, leading to the closing of the output gap by the end of the projection horizon.
According to our estimates, the crisis will have a downward impact on potential activity, although the losses should be considerably mitigated by the monetary and fiscal measures adopted to support businesses both during and after the lockdowns. In the medium term, the drop in productive investment should lead to a reduction in the stock of productive capital; moreover, the job losses and disruption of global supply chains should have a negative impact on market sector productivity. As a result, potential growth should fall to around half a per cent in 2020 and 2021, before increasing slightly to around 1% by 2023. It is then expected to remain close to its long-term rate, which is estimated at slightly over 1%. The loss of potential GDP compared to our pre-crisis projections is thus expected to be around 2% in 2023.
Box : Alternative scenarios show different ways the pandemic could evolve, with very different impacts on the pace of economic recovery (click to unroll)
Due to the Covid-19 epidemic, our baseline scenario is subject to a much higher degree of uncertainty than under normal circumstances. This box therefore presents two alternative scenarios to our baseline projection, reflecting “mild” and “severe” dynamics in the pandemic’s development (see Chart 3). These alternative scenarios result from simulations performed using our FR-BDF projection model and constructed based on a combination of scenario-specific shocks to domestic and foreign demand, and alternative assumptions for certain financial variables. As was the case in June, these scenarios were developed within the framework of the Eurosystem’s preparation of its joint projections in December 2020 and are thus based on a coordinated approach, with the European Central Bank and national central banks publishing analogous scenarios.
In the mild scenario, we assume that the public health situation worldwide improves rapidly and markedly from the beginning of 2021. We also assume that an effective medical solution, such as a marketed vaccine, is put in place in early 2021 and is deployed fairly rapidly and extensively throughout the year. This scenario would lead to a sharp recovery in activity both in France and internationally. In response to the success of the public health measures, which would lead to a more rapid lifting of restrictions and reduced uncertainty, households would quickly and significantly increase their spending and thus bring down the household saving ratio, while the outlook for investment would brighten considerably. Thus, after deteriorating at the end of 2020, the economy would follow a “W-shaped” recovery and GDP would return to its pre-crisis level by the end of 2021. After a still sharp decline in 2020, the rebound in 2021 and 2022 would be robust at 7% and then 5% (see table below). In this scenario, we consider that the pandemic would not have a lasting impact on potential GDP and inflation would pick up a little more markedly from its very low level in 2020, rising to 1.2% at the end of 2023. Consequently, government debt would slowly drop back over the projection horizon to a level close to 110% of GDP in 2023.
In the severe scenario, we assume that the virus continues to circulate in France and around the world with a high degree of virulence in 2021 and 2022. Strict public health measures continue to be needed, some of which remain in place until the beginning of 2023. Despite initial successes on vaccines, we assume that their widespread deployment is ultimately slow and does not reach the coverage required before the beginning of 2023. After the sharp fall in GDP in 2020, repeated or extended lockdowns and heightened economic uncertainty would continue to weigh on household and business spending and would lead to a further deterioration during the first half of 2021. In addition, with businesses failing in greater numbers in 2021, the economy’s capacity to recover would be impeded, leading to a significant loss of potential GDP. In this scenario, activity would not pick up in 2021 and would decline by 1%. Furthermore, the recovery in 2022 would continue to be gradual and restricted by long-term disruptions in the value chains and a high degree of economic uncertainty. Due to the very lasting reduction in economic activity, inflation would become negative in 2021 and 2022. Deflationary pressures would win out despite downward nominal wage and price rigidities. Consequently, government debt would exceed 130% of GDP as from 2021 and would rise sharply to a level close to 140% of GDP in 2023.
Compared to our June projections, our two alternative scenarios adhere a little more closely to our baseline scenario. This is linked to reduced uncertainty related to the economy’s short-term dynamics, which reflects a learning process since last May with regard to the understanding of the effects of public health measures on the economy, at least in the short term.
The various aid mechanisms put in place by the government should significantly cushion the income losses for private agents, and especially for households
The shock to activity in 2020 and the fact that the recovery will then be very gradual translate into a marked loss in national income. Economic agents (households, firms, general government) have different degrees of direct exposure to the shock and, in the case of private agents, the impact has been significantly cushioned by public aid mechanisms (short-time work, solidarity fund, etc.). According to our projections, general government should absorb 62% of the loss of gross national disposable income in 2020 (see Chart 6), which is much higher than its usual share of this income (25%). In parallel, households’ share of the losses should remain moderate in 2020 at 11%, well below their normal share of total national income, which is around 60%. The share incurred by households should increase slightly over time, especially as the labour market adjusts, but should still remain well below their usual weight in national income.
The extent of the government cushioning mechanisms should keep household purchasing power resilient in 2020. Despite the steep fall in activity, it is in fact expected to increase slightly, rising by an average of 0.3% per capita (see Chart 7). Of course, these aggregated figures mask considerable differences in circumstances depending on the household category, especially as the labour market adjustment should initially be concentrated in the most precarious forms of work. The lasting downturn in the labour market should also prolong the effects of the crisis on purchasing power per capita, which is only expected to see limited growth in 2021 (+0.3%), before rising more sharply as of 2022 (+1.4% in 2022, and +1.3% in 2023).
Household consumption should continue to be limited by the public health situation in 2021, but should then grow very sharply in 2022, depending on the evolution of the saving ratio
After rebounding significantly in the third quarter of 2020, household consumption is likely to be strongly curtailed again by the curfew imposed as of mid-October and, above all, by the lockdown as of end-October. In the fourth quarter of 2020, it is projected to be around 10% lower than in the same period in 2019. The deterioration in household consumption should be less severe than during the first lockdown (–17% in the second quarter compared with its pre-crisis level). However, it should still outstrip the fall in GDP (see Chart 8) as this second lockdown has weighed more specifically on shops and household services, in contrast with the restrictions during the spring. As in spring 2020, this forced reduction in household consumption should translate into a new peak in the saving ratio in the fourth quarter (see Chart 9). At the end of 2020, the additional financial savings accumulated by households, calculated as the deviation from a trend scenario, are expected to amount to EUR 130 billion.
Over 2021, the recovery in household consumption will be strongly contingent on the evolution of the health situation. The gradual improvement in the latter should allow household spending to rise again sharply, especially from the end of the year. In parallel, the saving ratio should decline progressively, although it should still be above pre-crisis levels at the end of the year (see Chart 9).
In 2022 and 2023,with the widespread roll-out of vaccines, the trajectory of consumption should be shaped by two factors other than income: on the one hand, whether households choose to build up precautionary savings in the face of the lasting deterioration in the labour market; and on the other, the extent to which they consume the excess savings accumulated in 2020 and 2021. These factors will have opposite effects on consumption, and we have chosen to assume that the household saving ratio will temporarily fall fairly sharply in 2022, before coming back up again to close to its end-2019 level by end-2023. This should lead to a temporary spike in consumption in 2022, which will be a major contributor to the strong GDP growth, and household consumption should only return to a pace of growth more in line with GDP in 2023 (see Chart 8). How household saving behaviour evolves is thus a major risk to our projection.
Business investment has fallen less than feared so far, and should be supported by the recovery plan
Although it has indeed declined, business investment has so far held up better than expected, falling to a lesser extent than GDP between end-2019 and the third quarter of 2020 (as often happens during a crisis).
In 2021 and 2022 it should be supported by the government stimulus measures and continuing favourable financing conditions; however, it is still only expected to recover gradually (see Chart 10) as the outlook for activity improves.
Non-financial corporations’ investment ratio (investment as a share of their value added) is expected to remain above its end-2019 level over the entire projection horizon, although the robust pace of growth seen before the crisis is projected to end. Due to the impact of the crisis on NFC income levels, their self-financing ratio is expected to decline by 10 percentage points over 2020 and 2021 and should remain below its 2019 level for the remainder of the projection horizon.
Unlike consumption, household investment should only fall to a limited extent in the fourth quarter of 2020, thanks to the continuation of activity in the construction sector. At the end of 2020, it should be around 11% below its pre-crisis level after the brutal slump of nearly 30% in the second quarter. However, household investment should then only recover gradually, as suggested by the sharp fall in building permit applications since last spring, and in the medium term by the deterioration in the labour market. As a result, the household investment ratio (investment as a share of household income) is not expected to return to the level observed between 2017 and 2019 before 2023.
Despite the recovery in world demand, France’s export-to-import ratio is expected to remain lastingly impaired
French exports suffered a major shock in the first half of 2020, dropping by 30% in the second quarter compared with their pre-crisis level (see Chart 11). The third quarter saw a marked rebound, however, and this trend should continue over the end of the year, despite the global health situation. The recovery should be supported by rising foreign demand for French goods and services, as manufacturing and business service activities have been allowed to remain open despite the second lockdown. The upward revision to extra-euro area foreign demand compared with our September forecast stems from the faster than expected recovery in Asia, which should help to offset the short-term impact of the resurgence of the epidemic in Europe on intra-euro area demand.
In the medium term, however, exports are expected to remain impaired as the ongoing restrictions on international travel continue to weigh on some of France’s stronghold sectors – notably tourism and aeronautics. French exports are only anticipated to return to their pre-crisis level in the second half of 2023.
Imports should drop sharply in 2020, but to a much lesser extent than exports, resulting in a marked deterioration in France’s export-to-import ratio (see Chart 12). Imports should be supported by the relative resilience of domestic demand. However, with exports only expected to rebound partially by the end of 2023, the export-to-import ratio is then expected to stabilise at this low level.
The unemployment rate should reach a peak of close to 11% in the first quarter of 2021 before declining markedly towards 9%
The first half of 2020 saw a sharp drop in employment with the disappearance of around 850,000 jobs (as measured in the quarterly national accounts). This 3% fall is nonetheless small compared to the size of the shock to activity, thanks to the short-time work schemes put in place by the government. Employment then rebounded sharply in the third quarter. There is a strong risk, however, that this rebound will prove temporary given the more difficult context over the end of the year. Indeed, the third quarter reflects sharp up-ticks in temporary work and probably in the number of jobs in certain personal and household services such as catering, accommodation and leisure. In theory, then, the reintroduction of tight restrictions in these sectors should trigger a sharp downturn in employment in the fourth quarter of 2020 and first quarter of 2021. Employment is then expected to bottom out in the first half of 2021 and should subsequently start to recover.
The unemployment rate (see Chart 13) is projected to reach a peak of 10.9% in the first quarter of 2021, before declining progressively to 9.1% at end-2022. Nonetheless, our projection for the unemployment rate remains uncertain due to difficulties in predicting the behavioural responses of certain sectors in the current context. It is notably based on the assumption that the return of the labour force to near pre-crisis levels, which began in the third quarter 2020 after the steep contraction in the first half, will be confirmed over the second half of 2021.
Given the deteriorated macroeconomic environment, inflation should only strengthen gradually and is seen remaining low over the entire projection horizon
Inflation as measured by the Harmonised Index of Consumer Prices (HICP) has fallen sharply over the course of the year, from 1.7% at the start of the year to 0.1% in October and 0.2% in November (preliminary flash estimates). This decline stems largely from energy prices which slumped in the first quarter of 2020 and have since only partially recovered. It can also be explained by the lasting slowdown in service prices, caused mainly by the shocks to the transportation (notably air transport) and accommodation sectors. Conversely, an acceleration in food prices was observed during the two lockdowns, and the cost of implementing the new health measures may have led to occasional price hikes in services such as hairdressing, vehicle repair and catering. However, these effects have proved temporary and have had no lasting impact on the general path of inflation.
The downturn in activity and weak growth in input prices should continue to weigh on French inflation over the coming quarters. Headline HICP inflation is expected to reach 0.5% in annual average terms in 2020 (after 1.3% in 2019), and should be around zero at the end of 2020 and start of 2021. In 2021 it should again average 0.5% for the year. Then in 2022 and 2023 it is projected to strengthen slightly, first mechanically due to the base effect of the sharp falls in energy and service prices in 2020, and then more structurally as the recovery in economic activity drives up the non-energy, non-food component. Inflation should nonetheless remain weak over the entire projection horizon, reaching 1.1% at the end of 2023.
HICP excluding energy and food is seen following a similar trajectory to headline HICP, but within a narrower range: it should average 0.6% for 2020 then 0.5% in 2021, before recovering gradually to 0.9% at end-2023.
The unprecedented emergency support measures combined with the slowdown in economic activity are causing a sharp and lasting deterioration in the government deficit and debt
The extent of the health crisis has forced the government to use its fiscal policy tools by rolling out emergency support measures and announcing a recovery plan, financed in part by transfers from the European Union (these projections take into account the budget law for 2021 and the fourth amended budget law for 2020; they also incorporate the France Relance recovery plan announced by the government, and are based on the assumption that the European recovery plan will be implemented in 2021). The implementation of emergency measures and sharp contraction in GDP in 2020 have led to a marked deterioration in government finances over the year, which should only be partially reabsorbed over the projection horizon. In line with the alternative macroeconomic scenarios, we have also constructed two alternative scenarios for government finances which are described in the box (see above).
In 2020, the government deficit is expected to widen to over 9% of GDP from 3% in 2019 (2.1% excluding the double-counting of the CICE and reduction in social security contributions), due to the slump in GDP and the exceptional measures introduced to tackle the health crisis. On the one hand, government expenditure should be driven sharply upwards by the emergency measures, especially the exceptional short-time work scheme (EUR 30 billion according to our estimates), the additional health expenditure (EUR 12 billion excluding the Segur healthcare spending plan) and the reinforced solidarity fund for small businesses (EUR 19 billion). On the other hand, tax and social security receipts are expected to decline in 2020 due to the sharp contraction in activity and, to a lesser extent, the exemptions to social security contributions granted as part of the emergency measures (EUR 8 billion), and the measures introduced previously under the initial 2020 budget law. The widening of the government deficit and contraction in GDP, combined with other measures that have a neutral impact on the deficit but affect the debt (notably the state’s acquisition of financial stakes in strategic companies), should drive government debt up sharply to around 116% of GDP in 2020, from 98% in 2019.
In 2021-23, based on unchanged legislation, the government deficit should narrow progressively, helped by the return to economic growth, the end of the exceptional measures and the receipt of EU funds, despite the increase of the recovery plan. The deficit is anticipated to exceed 7% of GDP in 2021 and 4% in 2023. The roll-out of the recovery plan should keep government expenditure at a high level in 2021 and 2022, although it should decrease progressively as the measures linked to the health crisis are wound down. The ratio of government spending to GDP (excluding tax credits) should still be above 55% in 2023, compared with 54% in 2019. Tax and social security receipts are expected to rebound with the return to economic growth, but will still be curtailed by the removal of local business taxes under the recovery plan, and the tax cuts voted into law before the crisis. The tax ratio (total tax and social security receipts as a share of GDP) is projected to fall to slightly under 44% of GDP in 2023 from 44.1% in 2019.
In the absence of any new fiscal consolidation measures, in 2023 the government deficit is expected to remain well above the level anticipated before the health crisis. It should also exceed the level that would allow the government debt ratio to remain stable. Despite the favourable impact of low interest rates, which are below the rate of growth (placing downward pressure on government debt), France should thus be unable to lower its debt ratio, which will instead continue rising to close to 120% in 2023.
While the evolution of the pandemic remains the main risk to our projection, other factors are also sources of uncertainty
Aside from the uncertainty linked to the evolution of the epidemic in France and in our trading partner countries, as described in the box, there are a number of other risks to our projection.
With regard to Brexit, in line with the Eurosystem and European Commission’s projections for the euro area, our projection is based on the standard assumption that there will be no deal between the United Kingdom and European Union, but that this will not lead to financial tensions.
On a domestic level, the pace at which households spend the considerable additional savings accumulated in 2020 will also play a major role in determining the path of household consumption. It could be higher or rise more rapidly than in our baseline forecast or, conversely, households could choose to build up additional precautionary savings owing to the fragile labour market.
With regard to inflation, while the evolution of the macroeconomic environment suggests prices will remain sluggish, inflation could strengthen more markedly as the macroeconomic situation improves, and certain sectors could try to restore their profitability. Inflation could also be temporarily higher if firms incur additional costs linked to health and safety measures, and at some point pass these on to final prices.